The exponential washoff model is currently widely used within software such as PC-SWMM to estimate within-event and total loading of different non-point stormwater pollutants in urban watersheds. Data from a previous study of particulate and dissolved stormwater pollutants from 126 events at 12 sites in Singapore were used for event based model parameter calibration. Best-fitting exponential washoff model parameters were determined using a Monte Carlo method with no underlying assumptions regarding surface buildup. Values for washoff coefficent parameter c3 and ‘intial mass on surface’ parameter Bini were found to be log-normally distributed, and varied by up to 3 orders of magnitude for the same pollutant at the same site. Washoff exponent parameter c4 however was found to be normally distributed with an average value of close to 1.0 for all pollutants. Model accuracy and correlations between model parameters and rainfall variables were found to be highest for particulate pollutant TSS, and weakest for dissolved pollutant PO4. The strongest of these correlations were between model parameters c3 and Bini with total rainfall depth. No correlation was found between antecedent dry period and any model parameters for any pollutant, in line with earlier studies which have also raised doubt around the importance of buildup on the surface when modelling pollutant washoff, particularly in tropical catchments.